1. Field of Invention
The present invention relates to telecommunications call processing. More specifically, it relates to processing of a vanity telephone number dialed by a caller with a conventional telephone, so as to access a national virtual telephone number database to provide benefits, such as improved connection efficiency, selected services or products, to the caller, the servicing location(s) associated with the vanity number dialed and/or the vanity number advertiser.
2. Description of the Related Technology
Traditionally, entities with multiple employees, departments and/or locations, such as businesses and government agencies, have provided their customers with multiple telephone number points of contact, with usually at least one telephone number for each employee, department and location. This has placed a major burden on customers and prospective customers to find, remember, dial and be connected to the correct intra-entity telephone number for the services desired. It also has created cost and administrative burdens on these entities to publish and advertise multiple telephone numbers.
In the new world of electronic commerce, many such entities have started advertising “one number”, vanity telephone numbers as their primary customer contact point. These vanity numbers are usually national 10 digit numbers starting with area codes such as “800,” “888,” or “900”, local 7 digit numbers starting with an exchange such as “555” and “950” or special purpose three digit numbers like “311”, “411” or “911”. These numbers are usually easy to remember, such as 1-800-FLORIST. Unlike regular telephone calls with only two participants, vanity telephone number calls can have three participants, recipients, or beneficiaries:    1. The Vanity Number Advertiser    2. The Caller or Consumer    3. The Servicing Location(s)
Based on the increased volume of calls to these vanity numbers and customer demands for 24 hour support during seven days a week, reduced telephone busy signals and shorter hold times, vanity number advertisers have begun answering these calls with a new technology called Voice Response Units (VRU), also known as Interactive Voice Response (IVR).
Currently, there are over 50 manufacturers of VRUs. The commercialization of the VRU and changes in advertising practices has also spawned large numbers of new VRU applications from product manufacturers. Products may be advertised by an infomercial showing an “800” number to call so that a consumer may obtain a list of nearby dealers and/or a product brochure. The 800 number is answered by a VRU which requests the caller to record their name and address. This partially automates the call process, but requires large amounts of disk storage to store the caller provided recorded voice information and creates a large amount of post call work for the advertiser. For example, the advertiser must listen to, understand, transcribe the caller's name and address, certify the address by use of a United States Postal Service (USPS) coding accuracy support system (CASS), manually compile a list of nearby dealers and mail the information packet to the caller's address. These inefficiencies have created a need to further automate VRU applications. This is accomplished through what is now called intelligent call processing technology.
In this context, automated intelligent call processing (ICP) is defined as the capture of network provided data, such as automatic number identification (ANI) and dialed number identification service (DNIS), and caller-provided data, such as data entered by Dual Tone Multi-Frequency (DTMF) through a Touchtone telephone key pad or the caller speaking through the telephone to a VRU. ICP also involves the VRU accessing external databases that can decipher, validate, process and fulfill the caller's request by playing pre-recorded messages, creating call specific messages and speaking them to the caller, storing call captured information that can be accessed by or forwarded to the caller, servicing location or vanity advertiser, and/or automatically routing and connecting the caller to the servicing location or department. Semi-automated intelligent call processing is characterized by automating components of the call through intelligent call processing, but having some portion of the request still requiring live operator support during the call.
There are three primary components to an intelligent call processing system:                the network: the system level hardware and software that provides the platform for intra- and inter-system and participant communications;        the information retrieval, processing and storage: the databases and processing algorithms that provide the network application with the information required to fulfill the request; and        the applications: the processes that process and fulfill the request(s) of the caller, the servicing location and/or the vanity advertiser by utilizing the network and the retrieved, processed and stored information.The Network        
The VRU is the device that can be used to replace the network operator and/or the answering party. Early primitive, non-integrated ancestors to the VRU are the caller ID box and the answering machine. Current state-of-the-art VRUs are programmable devices that not only capture and process network provided data but also accurately translate caller spoken numbers and words into textual or binary data, and convert digital text in the form of words and sentences into speech that is understandable by most callers. The VRU capabilities in these areas are continuing to rapidly improve. The last remaining obstacle to VRU automation is immediate access to more information. This required better network access to network and remote databases and a way to associate the digital data stored in these databases with network provided data, such as ANI and DNIS, and caller provided telephone input in the form of sound: voice or DTMF accurately translated into digital data.
The computer network portion of this problem has been addressed with faster 32 bit and 64 bit processors, vast amounts of cheap RAM and disk storage, new levels of Computer Telephone Integration (CTI) and advances in computer wide area networking that provides real time access to many different databases stored on different computer systems physically located in different parts of the country. This is demonstrated in part by a variety of consumer computer-interface applications supported by computer network services, such as CompuServe®, America Online®, Microsoft Network™ and the Internet.
There are nearly 200 million access points in the national telephone network, which is many times the current number of access points for all of the computer networks combined. The major limitation of the telecommunications voice network is that other than the limited amount of network provided data and voice, the only widely supported communications means is another form of sound, i.e., DTMF, which is a very primitive way of achieving one-way communication. Voice recognition has improved tremendously over the last few years, but is still a long way from being able to translate the words spoken by millions of people with different voices and accents into digital text words with 100% accuracy.
A few access points have videophones that support both sound and video in both send and receive modes. The technology has been around for many years to convert digital text data into video, and digital raster data in the form of maps and pictures into video, and transmit it over the national telecommunications network. There is also primitive technology available to scan and translate video images in the form of hand-written messages and typed characters, words and sentences into digital data, such as the ASCII character set. Today, none of the VRU manufacturers provide either of these capabilities with their current products. As videophones become more common in use, the existing technology to translate digital data into the form of a video image and transmit it to the caller will likely become a standard feature in all next generation VRUs.
A few access points also have computers with modems, speakers, microphones and telephone emulation software, such as Microsoft Phone. There is potential to have the computer translate on-screen typed text into DTMF tones using a more robust DTMF coding scheme and to have this translated back into digital text at the VRU. However, current VRUs do not have this capability.
Information Retrieval, Processing and Storage
Currently, VRUs have no caller-friendly capability to accurately translate caller voice or DTMF input into complex digital database access keys. Consequently, VRU database access has been limited to databases indexed by a simple numeric key. These include pre-recorded messages and internal client customer databases indexed by customer ID. The ID is usually in the form of a telephone number, account number, policy number, order number or other numeric data that is provided by the network, can be entered by DTMF, or accurately translated into digital data by a VRU using current voice recognition technology. This method works for applications with existing customers who know their customer ID. However, for new customers, new businesses or new applications that service different target markets, these internal databases are either too sparse in coverage or do not contain the required information.
On the other hand, there are many frequently updated national databases that have not been accessible by VRUs using network provided data or caller provided telephone input. These include:                The USPS address coding guide.        The US Census Bureau's TIGER (Topographical Integrated Geographic Encoded Record) and 1990 census data files.        Geographic and spatial files from Geographic Data Technology, Inc. (GDT) and ETAK®, such as ZIP+4 to latitude and longitude, ZIP+4 to census block, ZIP Code and census boundary, and enhanced TIGER files.        Household and individual databases from Polk, First Data Resources (FDR), Metromail and the big three credit bureaus: Equifax, Trans Union and TRW.        Property databases from TransAmerica, TRW Ready Data and ACXIOM DATAQUICK.        Updated census data files and geodemographic databases from Claritas, Equifax National Decision Systems, Urban Decision Systems (UDS), CACI and Strategic Mapping, Inc. (SMI).        Business and government location databases from American Business Information® (ABI), DUNS, ProCD and Database American.        Business financial databases from DUNS and TRW.        Hundreds of private company, state and local government and regional files of various types.        
All the above databases have one or more of the following limitations that has previously restricted them from being used in VRU applications:                They do not contain a telephone number field.        They contain a telephone number field but a high percentage of records have missing telephone numbers, have out of date telephone numbers or have a very limited amount of data associated with the telephone number.        They do not share a common access key that the caller knows, is willing to provide and can easily communicate to a VRU.        
The missing link in making all the above data available in real-time to VRU applications is creating a standardized, precise and universal database linkage key that can be assigned to all the United States telephone numbers and all of the above mentioned databases. This key needs to act as a direct and/or translator linkage mechanism between the telephone number and databases for spatial, geographic, USPS address, household, individual, business location, government location, business financial, property and client service locations with service areas of any defined geographic size and shape. Since the most common trait shared among the above mentioned databases is their geographic/spatial location, definition and/or relationship, the most logical solution would be a universal hierarchical geographic/spatial linkage key, “Spatial Key”. Utilizing the Spatial Key to create a virtual telephone number database would make it practical to automate many VRU applications that provide the caller with information, connect the caller with a servicing location and/or capture or retrieve caller related information to assist the vanity advertiser and/or the servicing location in providing better during call and post-call service to the caller.
Applicant is not aware of any product or method that uses a single key to create a virtual telephone number database by linking to many different and seemingly unrelated databases for supporting multiple applications. Savage et al. (U.S. Pat. No. 4,954,958) associates the 10-digit telephone number with an address-indexed street network database to provide directions over a telecommunications network to a caller. Savage uses two 10 digit telephone numbers input by the caller to provide directions from point A corresponding to the location of the first telephone number to point B corresponding to the location of the second telephone number.
As a telephone number to address translation mechanism, the Savage system uses the American Business List (ABL) file which is compiled from the national yellow pages. The ABL file contains approximately 10 million unique business telephone numbers and was originally created for use as a direct marketing database and a national business directory assistance database. The Savage system indexes each 10-digit telephone number into the ABL File to retrieve a business name and a raw address for each end point. In the telecommunications and direct marketing industries, this well-known process of starting with a phone number and looking up a name and address from a directory database is called a reverse directory search. The Savage system uses the raw addresses retrieved by this process as a linkage mechanism to what is referred to as a geodata digitized mapping database from MapInfo®. The source of the MapInfo database most likely is the Census Bureau Geographic Base File-Dual Independent Measurement Encoded (GBF-DIME), which is the predecessor to the TIGER files.
There are many technical issues associated with using a raw, non-standardized and free-formatted address which is composed of a street number, street pre-direction, street name, street type, street post-direction, city name and state as a linkage means between two databases compiled from different sources. These issues include: field size, address formatting and parsing, upper case and lower case, abbreviations, alternate names, alternate spellings (First vs. 1st), missing components and the source of city name. For example, Highway 101, PC HWY, PCH, Pacific Coast Hwy, First Street and 1st St. are all valid alternate street names and types for 1st St. in Encinitas, Calif. This large number of address permutations requires very sophisticated address parsing, standardizing, sorting, matching and scoring algorithms to correctly match raw addresses from two independent databases.
The Savage system does not address the above issues in matching the two raw ABL retrieved addresses to their corresponding two raw addresses on the preferred MapInfo digitized mapping database. The Savage description of the address matching embodiment is: “the central processor will retrieve from the geodata digitized mapping database the routing data correlated to the geographic location addresses”. What is needed is a simple, accurate and definable way (such as a Spatial Key) to precisely hierarchically code the address associated with a telephone number and use it as a hierarchical match key to retrieve matching data from other databases coded with all or part of the same hierarchical match key.
In addition, the Savage system does not provide any automated means to determine a servicing location nearby the caller. The caller must know and input the telephone number of the desired service location to get directions. This also eliminates the possibility of providing directions to service locations, such as drop boxes and automatic teller machines (ATMs) that do not have telephones.
Riskin (U.S. Pat. No. 4,757,267) uses the first six digits of the caller's telephone number to select a nearby serving location by performing an on-the-fly calculation to determine the nearness relationship. However, none of the databases mentioned above are accessible by Riskin's process because the first six digits of the telephone number do not provide enough precision to identify the housing or business unit location of the caller.
There are also two previous systems that use a client-specific Caller Telephone Number To a Service Location Telephone Number table as a means of connecting a caller to a servicing location. Cotter (U.S. Pat. No. 4,797,818) describes a manually intensive process for building and maintaining this table. Wegrzynowicz (U.S. Pat. No. 5,136,636) only references the table as a system component that is built and maintained by the client, but does not describe how the client performs this function.
Neither Savage, Riskin, Cotter, nor Wegrzynowicz use a linkage process similar to the Spatial Key. Further, none of the prior systems mention using a single linkage mechanism as a means to link to multiple databases to support multiple applications.
Developing a Spatial Key
In developing a universal Spatial Key the following must be considered:    1. The stability and updateability of the key over time.    2. The ability of the key to be a unique housing, business and/or postal delivery unit identifier.    3. The geographic hierarchy and precision of the key.    4. The number and quality of updated commercial and public translation tables to and from the key.    5. The availability of tools for third parties to place the key on their files.    6. The ability to precisely associate the key to service locations with service areas of any geographic defined size and shape.    7. The ability of regulated telecommunications entities to code their files with the key and to pass the key outside the regulated portion of the network.
Based on the above considerations, there are four primary candidates for the key:                Most recent census block code        Latitude and Longitude        Telephone Number        USPS ZIP Code        
The other candidates, such as a voting precinct, are eliminated from discussion because of a lack of precision.
Most Recent Census Block Code
The Census block code is a hierarchical 15-digit Federal Information Processing Standard (FIPS) number that is updated once every 10 years in conjunction with the United States decennial census. It has the following seven level hierarchy:                2 digit state code        3 digit county code        4 digit tract code        2 digit tract suffix        1 digit block group code        2 digit block code        1 character block part code        
The critical limitation of using census block as the Spatial Key is it is not precise enough to act as a unique housing or business unit identifier.
Latitude and Longitude
Latitude and longitude are used in a spherical coordinate system to identify a point on the earth. Its stability in the United States is a function of the North American Datum (NAD) which was originally established by the United States Geological Survey (USGS) in 1927 and was updated in 1983. To use the latitude and longitude as a hierarchical key, the base 10 or binary digits of the latitude and longitude pair must be interleaved to form a single number. The result of this interleaving is generally referred to as a quadtree. Alternatively, the latitude and longitude pair may be combined and/or translated to form another identifier. When latitude and longitude are stored in millionths of degrees, the interleaving creates a nine level base 10 and a sixteen level binary hierarchical system with a mathematical precision of approximately plus or minus 4 inches.
This level of precision is supported by the US Department of Defense's implementation of Global Positioning Satellites (GPS) technology. However, the two primary commercial means by which latitudes and longitudes are assigned to a location, i.e., the TIGER files (NAD27) and commercial level GPS (NAD83), do not support this level of precision. For locations in California, the latitude and longitude coordinates vary by as much as 300 feet between NAD27 and NAD83. There is a mathematical relationship between NAD27 and NAD83, such that latitudes and longitudes can be converted back and forth.
In addition to the above precision issues, latitude and longitude would not make a good choice for a unique housing or business identifier because multi-story buildings require a third coordinate, i.e., elevation. Another limitation with latitude and longitude as a Spatial Key is it requires very specialized Geographic Information System (GIS) databases and knowledge to Spatial Key code. However, commercial level latitude and longitude has no equal when input into a GIS system using data from a single NAD that is indexed by quadtree in showing a relative location on a map with precision in the 30 to 100 foot range.
Telephone Number
The 10 digit telephone number appears to comprise a three level hierarchical system.                3 digit Numbering Plan Area (NPA) or area code        3 digit NXX, exchange or prefix        4 digit line number or suffix        
Currently, NPAs do not spatially overlap and, with two minor exceptions, do not cross state boundaries. However, there are current plans to create spatially overlapping NPAs in the future. This will require callers in these NPAs to always dial 10 digits. The next non-spatially overlapping level is not the NXX, but the central office (CO) or wire center (WC). Each CO supports one to a few NXXs. Usually over time, the line numbers associated with a NXX become randomly distributed across the locations of the households and businesses serviced by the CO. There are also NXXs, such as 555, 950 and those assigned to cellular phones and pagers, that have no specific geographic boundaries within the NPA. There are also non-spatial NPAs such as 800, 888 and 900. These above items could cause difficulties in an intelligent call processing system if the telephone number was used as the Spatial Key.
There are several additional deficiencies in using the telephone number as the Spatial Key. These include, for example, the situation of using the telephone number as a unique housing or business unit ID. However, there would be multiple IDs for housing units and businesses with multiple telephone numbers. This would lead to excessive complexity in the system due to the multiple IDs. The main negatives associated with using the telephone number as the Spatial Key are the difficulty of accurately coding other databases with a telephone number and the regulatory issues related to transporting telephone numbers obtained from regulated sources outside the regulated telecommunications network.
USPS ZIP Code
The ZIP Code at the 11 digit level is called the Delivery Point Code (DPC) or ZIP+6 and uniquely identifies an individual building, such as 123 N Main St. The DPC is the most precise geographic code presently supported by the USPS and can be used as a unique housing or business unit identifier for single unit structures. However, it cannot uniquely identify a housing or business unit in multiple unit buildings or firms.
The DPC is a geographic hierarchical numbering system of five levels defined as follows:                3 digit ZIP Code is called a Sectional Center.        5 digit ZIP Code is called a Post Office Service Area with a preferred USPS name called the last line name. This is the name shown on the last line of a mailing address. There are 3 special types of ZIP Codes. Two of these, “Fleet Post Office (FPO)/Armed Forces Post Office (APO)” and “PO Box only”, do not have precise spatial definitions, but can be linked to unique household equivalent mailing addresses.        7 digit ZIP Code identifies a geographic sector within a Post Office Service Area.        9 digit ZIP Code is called a ZIP+4 and is usually the geographic area of one side of a street within a single one hundred address range block. It is a unique household level identifier for most USPS' PO Box and RR addresses which usually do not have precise spatial definitions.        11 digit ZIP Code is called the Delivery Point Code or ZIP+6 and uniquely identifies a street number address, such as 123 N Main St. The street address is the most common USPS address and is a unique housing or business unit identifier for all single unit buildings with unique street addresses.Applications        
Historically, many high-demand telephone call processing applications have not been commercialized because of one or more technical or economic issues including: automated caller interface technology, integrating telephone and computer networks, and telephone number database validation, coverage, depth and linkages.
In addition, when the above issues are addressed, all known previous efforts in the technology have focused on a custom solution to a specific application, and not on an integrated system solution that meets multiple application needs and the needs of the caller, servicing location and/or vanity number advertiser.
Automated Applications
The following is a partial list of automated application examples that have not either been addressed by previous art or addressed with a highly customized individual solution. It would be desired for all these applications to be automated using a common architecture in which the caller dials a vanity number and the system captures the caller's 10 digit ANI and DNIS. The architecture would only require the caller to respond to application dependent system voice prompts and/or only input a telephone number, if a telephone number different from the ANI is required by the application.                Connecting a caller to a servicing location: The prior technology does not support service locations having service areas of any size and shape, nor situations where geographic precision is required. A solution is desired that provides these abilities in an integrated common architecture.        USPS address retrieval: This is presently addressed by having the caller record their name and address, which is later listened to by a person and transcribed. The transcribed address is then processed through CASS certified software for use in an existing customer database of addresses indexed by telephone number. What is desired is a way to use a caller provided telephone number to directly retrieve the CASS certified USPS address associated with the caller provided telephone number and, in applications requiring 100% accuracy, providing the caller a means to verify the retrieved address. In addition, in a post call process, the retrieved, verified and stored address and additional linked data is desired to be used by the vanity advertiser to mail to the caller, for example, a requested store coupon, menu, catalog or informational packet.        The VRU speaks the service location(s) name, address and/or micro directions (to the caller): Service location information is needed by the caller to mail, pickup and/or drop off something to a selected servicing location. The greatest need for micro-area directions to service location(s) is with service locations very small in size, such as Federal Express, UPS and USPS drop boxes, or ATMs located in large physical entities, such as shopping centers or multi-story buildings. A solution is desired that provides these abilities in an integrated common architecture.        The VRU speaks driveable street directions from the caller's location to the selected service location (to the caller). In addition, in a call parallel application, after transferring the call to the servicing location, the application retrieves the service location's FAX number from a Service Location Table and faxes to the service location the caller's telephone number, address and a map and/or directions from the service location to the caller location to assist the servicing location with delivery to caller. The Savage reference describes a application that requires the caller to input two telephone numbers, and the only benefactor to the Savage device is the caller. What is desired is a system that does not require the input of any telephone numbers, or at worst, only one telephone number is provided by the caller. In addition, services would be provided to the caller, servicing location and/or the vanity advertiser.        Eliminating servicing locations based on days and hours of operation and/or services offered: A solution is desired that provides these abilities in an integrated common architecture.        Caller profiling based on Census or geodemographic data: A system is desired, based on a caller's geodemographic code and product consumption rates, to only present product options to the caller that the caller is most likely to buy, or to route the call to an appropriate sales specialist based on the caller's profile.        Applications that require the caller's name and/or individual data such as product registration and insurance, loan or credit applications: What is desired is a way of linking a Spatial Key to a household database containing data, such as name of head of household, street address, number of children in the household and the names of other individuals living in or associated with the household. The system would speak these individual names and the caller would identify himself or herself. Then the system would link to individual data, such as date of birth, credit rating, and so forth, and provide it to the caller, servicing location, and/or vanity advertiser.        Business Location Data Retrieval: What is desired is a way of linking the caller's Spatial Key to a business database containing data, such as name of Business, SIC, Number of employees and DUNS number, which would link directly into the DUNS database for credit information.        Real Property Database Retrieval: What is desired is a way for a contractor, for example, before bidding on a job, to dial a vanity number that interfaces with an automated property database, enter the telephone number of the supposed residential property owner and verify the ownership, address, mortgage holder, and any outstanding liens on the property.Semi Automated Applications        
There are telephone call processing applications where operator decisions and/or assistance are required that can also benefit from a virtual telephone number database. The following are desired exemplary applications:                Address Lookup and verification by an operator taking a telephone order: In current telephone order systems, an operator key enters a customer's address and verifies the spelling with the caller. What is desired is a way for the caller's telephone number to be passed to the computer system to automatically retrieve the CASS certified address associated with the caller's telephone number and display it on the operator's visual display. The operator would then ask the caller for the address to which they want the order shipped. If the addresses match, the operator would not have to key enter it and verify the spelling with the caller. If the addresses are different, there is a high potential that the caller is trying to make a fraudulent order and the operator would ask additional questions required to make this determination.        Real Time Address to Spatial Key Coding and Spatial Key to Client Table with Off-Line Master Table update: What is desired is a way of continually updating a Master Table (Phone Number to Spatial Key table) that supports multiple clients and applications in the situation when a caller is trying to be connected to a servicing location and has provided a valid telephone number that is not in the Master Table.        “911” application: In a real time Public Health and Safety application, the caller places an emergency call to the emergency telephone number “911.” The “911” application costs the U.S. taxpayer several billion dollars each year, and is currently overloaded with non-emergency calls. What is needed is a more economical alternative system for non-emergency “911” calls that can alleviate the load from the current overworked system.        
A system and method that uses a single Spatial Key to create a virtual telephone number database by linking a caller's or caller provided telephone number to many different and seemingly unrelated databases for supporting multiple applications would be an advance in the industry. What is needed is an automated means to determine a servicing location nearby the caller, such that the caller does not need to know and input the telephone number of the desired service location to get directions or other desired information. This would facilitate providing directions to service locations, such as drop boxes and automatic teller machines (ATMs) that do not have telephones. Such a system would utilize all ten digits of the telephone number to provide enough precision to identify the housing or business unit location of the caller telephone number. What is desired is the integration of VRU technology with a CTI network and a virtual telephone number database to provides a way to support a host of applications that were not previously possible. Information benefits derived by the caller, the servicing location and the vanity advertiser would be made possible by retrieving information from a virtual telephone number database created through Spatial Key linkage technology. Thus, a single linkage mechanism as a way to link to multiple databases to support multiple applications is needed. A solution is desired that provides these abilities in an integrated common architecture.